CN112712730A

CN112712730A - Method for preparing a signal for controlling an at least partially automated vehicle

Info

Publication number: CN112712730A
Application number: CN202011154626.XA
Authority: CN
Inventors: H·梅伦兹
Original assignee: Robert Bosch GmbH
Current assignee: Robert Bosch GmbH
Priority date: 2019-10-24
Filing date: 2020-10-26
Publication date: 2021-04-27
Anticipated expiration: 2040-10-26
Also published as: US11981324B2; US20210122363A1; CN112712730B; DE102019216363A1

Abstract

The invention relates to a method for preparing a signal for controlling an at least partially automated vehicle (101), comprising the following steps: receiving environmental data; identifying, based on the environmental data, a vehicle (102) exiting a parking space that initiated or performed an exit parking space operation; predicting a shunting area (103) which is required by the vehicle (102) exiting the parking space for performing the exiting parking space operation; preparing a signal for operating the at least partially automated vehicle (101) based on the predicted shunting area (103).

Description

Method for preparing a signal for controlling an at least partially automated vehicle

Technical Field

The invention relates to a method for preparing a signal for operating an at least partially automated vehicle, to a device provided for carrying out the method, to a computer program for carrying out the method, and to a machine-readable storage medium on which such a computer program is stored.

Background

A method for monitoring and signaling traffic conditions in the surroundings of a vehicle is claimed in DE 102011087774 a 1.

Disclosure of Invention

The invention relates to a method for preparing a signal for operating an at least partially automated vehicle, comprising the following steps:

● reception of environmental data;

● identifying a vehicle exiting the parking space that initiated or performed an exit parking space maneuver based on the environmental data;

● prediction of the shunting area (Pr ä dizieren eines Rangierbereichs) which is required by the vehicle exiting the parking space for performing said exiting parking space maneuver;

● prepare for signals for maneuvering the at least partially automated vehicle based on the predicted shunting area.

An at least partially automated or partially automated vehicle is understood to mean a vehicle which is operated partially, highly or fully automatically. In particular, the automated vehicle may relate to a vehicle which is operated in an unmanned manner and which can be operated at least partially without human intervention. The vehicle may for example relate to a classic passenger car, a lorry, a mini-bus or a regular bus. The vehicle may additionally relate to other vehicle types, such as two-wheel or three-wheel vehicles.

The environmental data can be, for example, data relating to the environment of the automated vehicle recorded by means of sensors. As the sensor, it may be, for example, a sensor of a vehicle, such as a video sensor, a radar, a lidar sensor, and/or an ultrasonic sensor. It may also relate to acoustic data of the microphone. The sensor can also be mounted on another vehicle and/or on an infrastructure device.

Furthermore, the environmental data can also relate to already evaluated data of further traffic participants and/or external servers. For example, a number of data received from a server may have been aggregated and evaluated at the server.

The identification of the vehicle exiting the parking space is based on the environmental data. In this case, for example, different fields, in particular a driving area and a parking area, can be detected, and on the basis of this detection, the vehicles present in the detected fields can be classified into one category. If the vehicle is present on the parking area, the vehicle is identified as a parked vehicle, and if the vehicle is present on the driving area, the vehicle is identified as a non-parked vehicle. In this case, further influencing variables/characteristics can be taken into account, which are described in the exemplary embodiments. What can be determined from a number of features is: whether the parked vehicle is actually initiating or has performed an exit parking space maneuver. The indication that the parked vehicle is exiting the parking space or wants to exit the parking space is, for example, a headlight/headlight. Reversing light indicates an upcoming process of exiting the parking space, just like turning on dipped headlights. For example, the real-time speed of the identified vehicle can also be taken into account in the identification of the vehicle exiting the parking space. Additional features are introduced in additional embodiments and examples.

Based on the identification, a prediction is made of a shunting area that is required by the vehicle exiting the parking space for performing the exiting parking space operation. The prediction can be made, for example, by means of the predicted trajectory of the vehicle exiting the parking space, in combination with the geometric dimensions of the vehicle. The size of the vehicle can be realized, for example, by means of detected environmental data or by reading data of the identified vehicle model. The predicted trajectory can thus be transferred into the field covered by the vehicle/space covered by the vehicle, for example by means of the detected profile of the vehicle driving out of the parking space.

Preparing a signal for maneuvering the at least partially automated vehicle based on the predicted shunting area. The aim here is to avoid accident situations or blockage situations. For this purpose, the at least partially automated vehicle can be controlled, for example, in such a way that: avoiding intrusion into the predicted shunting area of the vehicle exiting the parking space until the exiting parking space operation is ended.

Instead of or in addition to: the identification of the vehicle exiting the parking space and/or the prediction of the shunting area can also be carried out by means of a vehicle-to-vehicle connection, i.e. a connection between two or more vehicles, or a vehicle-to-other object connection, i.e. between the vehicle and any other unit, such as a server, infrastructure or passerby, for example, via a vehicle mobile telephone. Here, what is conveyed by the vehicle itself that is exiting the parking space is, for example: the process of exiting the parking space is planned and the planned trajectory is communicated immediately forthcoming and, if necessary. Simultaneously or alternatively: it is also possible to transmit the field required for the parking space exiting operation. Furthermore, the planned duration of the exit parking space operation can be transmitted such that it is known to the at least partially automated vehicle that: during which time period the vehicle should not be driven on the desired field. It is also possible to receive signals from infrastructure equipment, for example improved lampposts with integrated cameras and integrated control devices. Such an infrastructure device can be designed for recognizing the course of exiting a parking space and for communicating with vehicles present in the vicinity at the infrastructure device itself. The identification by the infrastructure device can be carried out, for example, by means of an algorithm which is implemented in the infrastructure device and is trained by machine learning methods.

The signal for the actuation can be prepared, for example, at a control unit of the at least partially automated vehicle, so that an automated and thus very rapid reaction of the vehicle can also take place completely independently of the vehicle occupant or the driver of the vehicle.

The method can be implemented, for example, in a vehicle, on one or more vehicle control devices. Instead, it is possible to: the method can also be carried out outside the vehicle on a spatially separate server or on a control device in an infrastructure device, such as a traffic light.

The preparation can be effected by cable connection or wirelessly. If a transmission from the server or an external unit to the vehicle takes place, this transmission preferably takes place wirelessly, for example via a mobile communication network. When the method is carried out in a vehicle and the vehicle is operated, the transmission can also take place exclusively by cable connection.

The described method and system make this possible: the situation of a vehicle which is automatically driven, when automatically driven on a street with priority driving rights, being blocked by another traffic participant who is driving out of the parking space is avoided, since the required shunting area of the other traffic participant who is driving out of the parking space is not blocked. Furthermore, by means of this method, in the case of a reaction to the vehicle exiting the parking space, a comfort regulation, even an emergency regulation, of the driving behavior of the automated vehicle is possible.

The method furthermore provides the additional advantage that: on the one hand, the unpleasant driving situation of the vehicle, which is at least partially automated for the passengers, is avoided by reducing the uncomfortable braking operation; furthermore, the predictive capability for the preceding traffic situation is improved. In addition, not only is the user value increased, but the social acceptance of automated vehicles is also increased.

In a further embodiment of the method, the at least partially automated vehicle is controlled on the basis of the signal in such a way that: the vehicle does not intrude into the shunting area during an out-of-parking-space operation of the vehicle exiting the parking space.

Implementation of such behavior is possible through different operations to be performed. For example, the trajectory can be planned as follows: avoiding driving into the shunting area during the duration of the exiting parking space operation. For this purpose, for example, the speed, the acceleration or the planned lane selection can be adapted. For example, the speed can be reduced, a stopping point detected and a stop or a lane switch performed at the stopping point. For example, a stop line can be detected on the basis of the front shunting area in that: such that a front region (including a safety distance) of the at least partially automated vehicle does not touch the shunting area. From the current position of the at least partially automated vehicle to the acquired stop line, a trajectory can be calculated according to known methods, which trajectory brings the at least partially automated vehicle to a comfortable stop. Instead of or in addition to: when the beginning exiting parking space process is identified with some uncertainty, the speed of the partially automated vehicle may be reduced upon identifying a vehicle exiting the parking space. This makes it possible to carry out the stopping process comfortably.

In a further embodiment of the method, the at least partially automated vehicle is operated in such a way that: such that the at least partially automated vehicle does not exceed a predefined safe distance to the shunting area or the shunting area is increased by a predefined safe distance during the exiting parking space operation.

In this case, either a stop-line is determined on the basis of the predicted shunting area, which is not allowed to be traversed by the at least partially automated vehicle, or an additional stop-line is determined, which is added to a predefined safety distance. This second stop line replaces the first stop line in this case and should not or should not be allowed to be driven over.

Instead of or in addition to: in the case where the shunting area has been predicted, additional safety distances can be taken into account together, so that the field required for the exit parking space operation becomes large, and therefore the following area also becomes large: the at least partially automated vehicle should not or should not be allowed to intrude into this area.

The additional safety distance increases the safety for the passengers of the automatically driven vehicle and also for the passengers of the further vehicle in the case of automatic operation.

In a further embodiment of the method, the identification of the vehicle exiting the parking space is carried out on the basis of the detection of at least one longitudinal and/or transverse parking space and the detection of the occupancy level of the parking space.

By the acquisition of a lateral or longitudinal parking space and the vehicles present in the parking space, possible candidates can be determined in advance, which can be considered as vehicles exiting the parking space. Based on this acquisition, the vehicle exiting the parking space can then be identified.

In a further embodiment of the method, the identification of the vehicle exiting the parking space is performed on the basis of an analysis of the state of the turn signal lights, brake lights, backup lights, dipped headlights and/or interior lighting of the vehicle exiting the parking space. The mentioned features indicate that: the detected vehicle not only relates to a vehicle that is being parked, but also relates to a vehicle that is expected to start or has started in a short time. In particular, reversing the light means: the identified vehicle will initiate travel of a rearwardly directed trajectory in a short period of time.

In a further embodiment of the method, the identification of the vehicle exiting the parking space is carried out on the basis of an acquisition of an orientation of the vehicle exiting the parking space relative to the parking space and/or relative to a driving lane driven by the at least partially automated vehicle, in particular on the basis of an acquisition of a temporal change of the orientation.

Both the parking space and the driving lane traveled by the at least partially automatically operated vehicle can be acquired, for example, by means of map data and/or received environment data.

The temporal change in the orientation represents the movement of the vehicle exiting the parking space, and a conclusion can be drawn about the planned trajectory of the vehicle or the planned travel operation of the vehicle. In general, orientation, in particular when identifying laterally parked/parking vehicles in the driving direction, is a reliable indicator for identifying an upcoming process of exiting a parking space.

In a further embodiment of the method, the acquisition of the position is based on the detection of a contour of the driving lane, of the parking space and/or of the vehicle exiting the parking space.

In a further embodiment of the method, the following are obtained: the process of exiting the parking space is then imminent when the position of the vehicle exiting the parking space exceeds a predefined angle with respect to a previously acquired or received contour of the parking space and/or of the driving lane.

The override can be implemented in particular during the time of monitoring the position of the vehicle exiting the parking space. In other words, the orientation of the vehicle exiting the parking space does not initially exceed the angle. However, during the observation of the vehicle exiting the parking space, it changes over time and eventually exceeds the predefined angle. The contour of the parking space or of the driving lane can be determined by means of the detected environmental data. Instead of or in addition to: the contour can also be received or read, for example from a high-precision map.

In a further embodiment of the method, the identification of the vehicle exiting the parking space is based on the acquisition of an angle of the orientation of the vehicle exiting the parking space relative to a previously acquired or received contour of the parking space and/or of the driving lane.

If, for example, a predefined angle is exceeded, this indicates that the vehicle exiting the parking space is actually performing an exit parking space maneuver or is planned to perform such an exit parking space maneuver.

In a further embodiment of the method, the prediction of the shunting area is based on an acquisition of a wheel base for the vehicle exiting the parking space, in particular on a steering circle derived from the wheel base.

In particular, a maximum steering circle can be derived on the basis of the acquisition of the wheel base. On the basis of this maximum turning circle, a prediction of the required field for the process of exiting the parking space can be carried out.

The wheel base and/or the size of the vehicle exiting the parking space can be achieved by means of a recall and/or classification (video, radar, lidar) of the stored information. This information can be evaluated directly by known classification methods and/or measured values from sensor data (Vermessung), for example.

In a further embodiment of the method, the prediction of the shunting area is based on the acquisition of available street widths.

In a further embodiment of the method, the prediction of the shunting area is based on an acquisition of the number of travel movements required for the vehicle exiting the parking space for performing the exiting parking space operation, in particular taking into account the acquired vehicle size of the vehicle exiting the parking space.

In order to obtain the required number of travel movements, empirical values can be assumed or different possibilities can be simulated. The feasible solution using the largest shunting area may be considered as the final shunting area. In this way it is ensured that: automatically operating vehicles are prevented from driving into the shunting area during the exiting parking space operation. The acquisition of the required driving movement can also be undertaken, for example, by an algorithm which is trained by means of machine learning methods.

In a further embodiment of the method, the prediction of the shunting area is based on the identification of protruding objects and the acquisition of available travel paths on the basis of these objects.

Raised objects can be understood as e.g. infrastructure devices (such as street lights, traffic signs or the like), further vehicles, curbs, and other objects (such as stones, walls, or the like).

All of the above mentioned features that can be taken into account for predicting the shunting area are used to cause the predicted shunting area to be improved. The better the shunting area is obtained, the less the risk of possible congestion-or safety-endangering traffic situations.

Furthermore, the method described herein creates an apparatus configured to: the steps of the method variants described here are executed, manipulated or carried out in corresponding devices. The object on which the invention is based can also be solved quickly and efficiently by means of this embodiment variant of the invention in the form of a device.

For this purpose, the device can have at least one computing unit for processing signals or data, at least one memory unit for storing signals or data, at least one interface to the sensor or the actuator for reading sensor signals from the sensor or for outputting data signals or control signals to the actuator, and/or at least one communication interface for reading or outputting data, which are embedded in a communication protocol. The computing unit may be, for example, a signal processor, a microcontroller, etc., wherein the memory unit can be a flash memory, an EEPROM or a magnetic memory unit. The communication interface can be designed to read or output data wirelessly and/or by wire, wherein the communication interface, which can read or output the wire data, can read or output the data from or into the respective data transmission line, for example electrically or optically.

In the present case, an apparatus is understood to be an electrical device which processes sensor signals and, depending on this, outputs control signals and/or data signals. The device can have an interface that can be configured according to hardware and/or software. In the case of a hardware-based configuration, the interface can be part of a so-called system ASIC, for example, which contains the most diverse functions of the device. However, it is also possible: the interface is a self-contained, integrated conversion circuit or is at least partially composed of discrete components. In the case of a configuration according to software, the interface can be a software module which is present, for example, on a microcontroller in addition to other software modules.

A computer program with a program code can also be advantageously stored on a machine-readable carrier or storage medium, for example on a semiconductor memory, a hard disk memory or an optical memory, and used to execute, implement and/or control the steps of the method according to one of the embodiments described above, in particular when the program product or program is executed on a computer or a device.

A machine-readable storage medium on which this computer program is stored is likewise claimed.

Drawings

Fig. 1 illustrates an exemplary traffic situation.

Fig. 2 shows a schematic process diagram.

Detailed Description

In the first exemplary embodiment, a vehicle 101 is present which is operated fully automatically in the traffic situation illustrated in fig. 1, i.e., at least at times, the vehicle 101 is operated without driver intervention. The automated vehicle 101 is equipped with a number of environmental sensors, among which are cameras, lidar sensors, radar sensors, ultrasonic sensors, and microphones. The vehicle 101 is operated by a central control unit on the basis of the data detected by the surroundings sensors. The automated vehicle 101 travels on a street 105, at the edge of which a parking space for parking the vehicle is present. A vehicle 102 exiting the parking space is to exit the parking space rearwardly from the parking space 106. The parking space exiting operation necessary for this purpose may be merely converted, because of the available space, into a multiple towing operation, which is illustrated in fig. 1 by the trajectory 107 traveled by the vehicle 102 exiting the parking space. For shunting (rangiren), the vehicle 102 leaving the parking space requires a likewise marked shunting field 103. Furthermore, three

possible lines

104a, 104b, 104c are indicated, which indicate the stop position of the automatically driven vehicle 101 or possible driving limits.

In order not to hinder the exiting vehicle 102 during its exiting parking space maneuver, which could cause a blockage situation, the method outlined in fig. 2 is executed in the automatically traveling vehicle 101, which method starts in step 201.

In step 202, environmental data are received from control devices of the automated vehicle 101. These environmental data come on the one hand from sensors of the vehicle itself; on the other hand, signals from other automated vehicles and correspondingly established infrastructure elements are additionally received. The determined object, its size, position and expected trajectory are particularly relevant here.

In step 203, the vehicle 102 that is exiting the parking space and is performing an exiting parking space operation is identified based on the received environment data. The different received and recorded intent characteristics of the vehicle 102 exiting the parking space are evaluated and weighed. On the one hand, variables of the dynamics of the vehicle, in particular the movement of the vehicle, are evaluated. In addition, the probability is measured whether the vehicle is expected to move in the near future and, if necessary, whether an exit from a parking space process is carried out. For this purpose, the intended features, such as the state of the headlights, the tire position, the engine noise and the presence of the vehicle occupants, are taken into account. Since in this embodiment the vehicle 102 exiting the parking space is already in the process of performing the exiting parking space operation and here the automatically traveling vehicle 102 has blocked the lane 105, the identification is primarily based on the position of the vehicle 102 exiting the parking space and the speed of this vehicle, as well as the trajectory that has been exited by the vehicle 102 exiting the parking space.

In step 204, a shunting area 103 is predicted, which is required for the vehicle 102 exiting the parking space to perform the exiting parking space operation. For this purpose, the size of the vehicle 102 exiting the parking space, i.e. the width and length of the vehicle, is determined on the basis of the detected environmental data. Furthermore, a trajectory 107 of the vehicle 102 exiting the parking space is predicted. Furthermore, the trajectory 107 is predicted based on the wheelbase of the vehicle 102 exiting the parking space, which is derived from the environmental data. The shunting area 103 is determined based on the predicted trajectory and the size of the vehicle 102 exiting the parking space.

In step 205, a signal for controlling the automated vehicle 101 is provided on the basis of the predicted shunting area 103. Based on the signals provided, the automated vehicle 101 is controlled in such a way that: during the duration of an exiting parking space operation of the exiting vehicle 102, the automated vehicle 101 is prevented from penetrating into the shunting area 103. For this reason, in this exemplary embodiment, the speed of the automated vehicle 101 is reduced, so that it is avoided that the vehicle 101 travels past the limit line 104c during the duration of the parking space exiting operation.

The method ends in step 206.

In another embodiment, the automatically running vehicle is in a situation comparable to the situation shown in fig. 1.

In order to prevent a vehicle 102 exiting the parking space from penetrating into the shunting area 103, three

different stop lines

104a, 104b, 104c are calculated. When the automated vehicle is parked at the height of line 104Aa, the vehicle 102 exiting the parking space may not be able to perform the planned exit parking space operation because the automated vehicle 101 will block the desired shunting site 103. To avoid this, a method comparable to the method outlined above is operated in the automated vehicle 101. However, in a departure from the method sequence shown above: the limit line 104b is first determined. When parked at this limit line 104b, the vehicle 102 exiting the parking space will be able to perform its exiting parking space operation unimpeded. However, the position is very restrictive in terms of size. A predefined safe distance is therefore planned in this embodiment and the limit line 104c is determined on the basis of this safe distance. In this exemplary embodiment, the automated vehicle 101 is controlled on the basis of the provided signals in such a way that: the vehicle does not travel beyond the limit line 104 c. For this purpose a stop operation is performed at the limit line 104 c.

In a further embodiment, the method has the following steps for providing a signal for operating an at least partially automated vehicle.

An exit parking space procedure for a vehicle exiting the parking space is first identified. An environment sensor detects the environment of the automated vehicle and makes this possible: the forward traffic situation is interpreted, predicted and further driving decisions are made for the automated vehicle taking into account the expanded map information (localization map and planning map). In addition, the situation with a front longitudinal parking space and, if applicable, a transverse parking space is detected and the occupancy level of the parking space is determined. It was additionally determined that: whether a vehicle is exiting the parking space from an occupied parking space. Particular attention is paid here to the recognition of the exit of a parking space from a lateral parking space, since a larger shunting area is required as a result of this.

Upon detection of a vehicle exiting a parking space, a shunting place required for performing an exiting parking space operation of the vehicle exiting the parking space is determined. The required shunting site is estimated here by a method in which: on the one hand, the wheelbase of the vehicle exiting the parking space is determined and, on the basis of this information, the maximum steering circle is derived; on the other hand, the available street width is determined. In addition, the available travel path is determined by determining other protruding objects.

With the aid of this information it can be deduced that: the vehicle exiting the parking space on the lane can exit the parking space for further driving in a certain number of driving movements. The shunting area that may be required is determined by means of this trajectory and the size of the vehicle. The track width and the size of the vehicle are determined here by classification algorithms (based on video data, radar data, and lidar data) and calls for stored information.

After the shunting region is determined, a stopping point is determined, beyond which the automated vehicle should not travel. Based on the shunting field in front, a stop line is determined for this purpose in such a way that: so that the front region of the automated vehicle, including the safety distance, does not touch the shunting region or intrude into this shunting region.

In order to ensure that the automated vehicle does not actually enter the shunting area, a trajectory for reaching the determined stopping point is determined and set. According to known methods, a trajectory is calculated from the current position of the automated vehicle to the stopping point, and when the trajectory is traveled, the automated vehicle stops comfortably before penetrating into the shunting area.

In a further embodiment, the speed of the automated vehicle is already reduced when the vehicle exiting the parking space is detected. This makes it possible to put the stopping process at the determined stopping point into practice comfortably.

Claims

1. Method for preparing a signal for handling an at least partially automated vehicle (101), comprising the steps of:

● receiving (202) environmental data;

● identifying (203) a vehicle (102) exiting the parking space initiating or performing an exit parking space operation based on the environmental data;

● predicting (204) a shunting area (103) which is required by the vehicle (102) exiting the parking space for performing the exiting parking space operation;

●, signals for operating the at least partially automated vehicle (101) are prepared (205) on the basis of the predicted shunting area (103).

2. The method of claim 1, wherein the at least partially automated vehicle (101) is steered based on the signal by: so that the at least partially automated vehicle does not intrude into the shunting area (103) during an exiting parking space operation of the vehicle (102) exiting the parking space.

3. The method according to claim 2, wherein the at least partially automated vehicle (101) is steered by: such that the at least partially automated vehicle does not exceed a predefined safety distance to the shunting area (103) during the exiting parking space operation, or the shunting area (103) increases by the magnitude of the predefined safety distance.

4. Method according to any one of the preceding claims, wherein the identification of the vehicle (102) exiting the parking space is carried out on the basis of the detection of at least one longitudinal and/or transverse parking space (106) and the acquisition of the occupancy level of the parking space.

5. The method according to any one of the preceding claims, wherein the identification of the vehicle (102) exiting the parking space is based on an analysis of the state of turn signal lights, brake lights, backup lights, low beam lights and/or interior lighting lights of the vehicle (102) exiting the parking space.

6. Method according to any one of the preceding claims, wherein the identification of the vehicle (102) exiting the parking space is carried out on the basis of an acquisition of an orientation of the vehicle (102) exiting the parking space relative to a parking space and/or relative to a driving lane driven by the at least partially automated vehicle (101), in particular on the basis of an acquisition of a temporal change of the orientation.

7. Method according to any one of the preceding claims, wherein the identification of the vehicle (102) exiting the parking space is based on the acquisition of the angle of the orientation of the vehicle (102) exiting the parking space relative to the previously acquired or received contour of the parking lot and/or of the driving lane.

8. Method according to any of the preceding claims, wherein the prediction of the shunting area (103) is based on an acquisition of a wheel base of the vehicle (103) exiting the parking space, in particular on a steering circle derived from said wheel base.

9. The method according to any one of the preceding claims, wherein the prediction of the shunting area (103) is based on an acquisition of available street widths.

10. Method according to any one of the preceding claims, wherein the prediction of the shunting area (103) is made on the basis of an acquisition of the number of driving movements required for the vehicle (102) exiting the parking space for performing the exiting parking space operation, in particular taking into account the acquired vehicle size of the vehicle (102) exiting the parking space.

11. The method according to any one of the preceding claims, wherein the prediction of the shunting area (103) is based on the identification of protruding objects and the acquisition of available travel channels on the basis of these objects.

12. An apparatus for performing the method of any one of claims 1-11.

13. A computer program comprising instructions which, when executed by a computer, cause it to perform the method according to any one of claims 1 to 11.

14. A machine-readable storage medium on which a computer program according to claim 13 is stored.